[RFE] Support multiple threads for netperf

[smalleni]

It would be nice to have the capability to run multiple threads of netperf to saturate high speed networking cards especially at smaller packet sizes.

[smalleni]

@jtaleric mentioned that implementing the super netperf script should help accomplish this: https://github.com/borkmann/stuff/blob/master/super_netperf

[jtaleric]

So, with https://raw.githubusercontent.com/borkmann/stuff/master/super_netperf we need to make modifications.

We are using the OMNI output. This script doesn't account for this, so we will need to fork, and make modifications.

[jtaleric]

https://github.com/jtaleric/tinker/blob/main/networking/super-netperf

[root@polaris /]# bash super.sh 5 -t TCP_STREAM -l 2 -- -k rt_latency,p99_latency,throughput,throughput_units
MIGRATED TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to localhost () port 0 AF_INET : demo
RT_LATENCY=-1.000
P99_LATENCY=32
THROUGHPUT=91326.96
THROUGHPUT_UNITS=10^6bits/s
[root@polaris /]# bash super.sh 5 -t TCP_RR -l 2 -- -k rt_latency,p99_latency,throughput,throughput_units
MIGRATED TCP REQUEST/RESPONSE TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to localhost () port 0 AF_INET : demo : first burst 0
RT_LATENCY=107.193
P99_LATENCY=63
THROUGHPUT=233339.35
THROUGHPUT_UNITS=Trans/s
[root@polaris /]# bash super.sh 5 -t TCP_CRR -l 2 -- -k rt_latency,p99_latency,throughput,throughput_units
MIGRATED TCP Connect/Request/Response TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to localhost () port 0 AF_INET : demo
RT_LATENCY=433.730
P99_LATENCY=263
THROUGHPUT=57688.43
THROUGHPUT_UNITS=Trans/s
[root@polaris /]# bash super.sh 1 -t TCP_STREAM -l 2 -- -k rt_latency,p99_latency,throughput,throughput_units
MIGRATED TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to localhost () port 0 AF_INET : demo
RT_LATENCY=-1.000
P99_LATENCY=11
THROUGHPUT=42291.62
THROUGHPUT_UNITS=10^6bits/s